Chatbot use cases – FAQ bot

Providing a smooth, enjoyable Q&A chatbot experience is key to your business success. Deploying an FAQ-based chatbot can not only up to your customer satisfaction levels but also increases your customer base due to higher conversion rates.

Building a chatbot for frequently asked questions (FAQ) helps businesses to engage customers 24×7 by answering their simple queries. It resolves queries in real time that reduces the number of support tickets and increases team productivity.

While creating the chatbot, it should be based on the business requirements. Go through certain steps to make a FAQ chatbot.

1. Define business objectives

The first and foremost step is to outline your business functions that need to be automated. What exactly will your chatbot do? Businesses build chatbots for websites for different purposes. They could be:

• To drive sales or collect leads by booking appointments, demos, sign-ups, etc. 
• Enhance online presence by engaging customers 24×7 when the support team is busy.
• Automate customer support with FAQs and deliver hybrid support (a bot will handle simple queries and transfer the complex ones to the live agents)
• Collect customer feedback to understand their satisfaction level and improve customer experience.

2. Identify the commonly asked questions

The business receives thousands of queries every day. You need to filter the frequently asked questions. Analyzing the common questions asked by customers is a great way to build an effective FAQ chatbot.

3. Categorize the FAQs 

Review the common queries and categorize them into different sub-categories like – sales, support, billing, etc. When you make different categories of common queries, it provides clarity in building a robust chatbot that aligns with your customer needs.

4. Create multiple variations of questions

Creating a number of options for common queries helps the chatbot to understand the intent of the user better. It trains the bot to identify the user intent and have effective communication with the users. 

If you only want your bot to answer questions you auto-populate on your site, then it won’t have to learn as much as it would if you want it to respond to user inquiries akin to a human agent. 

5. Give your bot a voice and personality

You can enhance your bot personality by giving a name and picture that complements your brand message. Customizing the bot adds a humanized touch and makes user interaction comfortable and friendly.

6. Build the conversational flow for FAQ chatbot

Defining the flow is a vital part of building your FAQ chatbot. Building the conversational flow is an exercise to help you organize your content and start thinking about the best ways to word the bot’s answers. 

The dialog flow goes into all the detail required to implememanized with the help of nodes and actions.

• A node refers to a specific scenario or a part that you want to execute the bot flow such as a welcome greeting or a product overview. 
• Actions are interactions that bots have with visitors. Actions can be combined with different elements to drive conversations forward in a pre-defined manner.

7. Create a balanced approach 

Most of the chatbots may not be that effective and customers look for chat support at some point. You can define at what particular stage, your customers can get a chance to connect with the human agent. It will create a balance between the bot and the live agents which can reduce shopping cart abandonment rate.

The other way to do the same is by defining fallback scenarios. It helps to identify the break-up points in the conversation and define an alternate solution that the bot can offer to the users. 

8. Measure your chatbot performance

While building a FAQ chatbot for your website, it is important to understand the key chatbot metrics and KPIs. They help to know how your bots are performing to achieve your goals.

In order to make your chatbot a successful one, it’s important to define your KPIs and measure them. Based on your business objectives, you can identify how you will be measuring your chatbot analytics.

The key KPIs to measure for FAQ bot are:

• No of queries resolved by the bot –  It evaluates the queries resolved by the chatbot successfully. 
• Response time – The KPI measures the total time taken by the bot to resolve the query.
• Total handover – It checks on how many out of total conversations were complex for the bot to understand and transferred to the live agent.

Note: You can measure the KPIs weekly or monthly in order to understand the bot performance and where you may need to fine-tune the bot flow to improve your KPI. It helps assess the effectiveness of your chatbot but gains insights into how your customers perceive your bot service. You can understand how far your customers are satisfied after interacting with the bot.

How to create a FAQ chatbot?

In order to create a FAQ chatbot, you need to follow the below steps.

1. Click on ‘Create New Bot’ under the ‘Bots’ option on the REVE Chat dashboard.

2. The second step is to set the profile of your FAQ bot. It provides you different things to customize the bot profile.

• BOT name – Setting up a bot name helps for internal business purposes. This name is not displayed to your customers. (For eg – FAQ or Sales bot).
• Display name –  As this name is displayed to your customers, you need to choose a name that matches your website branding. (For eg Annie or Amy). The displayed name makes customers feel that they are interacting with the human agents.
• Tagline – You can set a designation of the chatbot, which can also be displayed to visitors and customers.
• BOT Avatar –  It allows you to personalize your bot by choosing an avatar from the gallery or upload of your choice. 
• Enable channels –  REVE Chat offers two channels for bot communication. (i) Website (ii) Facebook. You can choose your preferred channel or both to deploy the chatbot for 24×7 customer assistance. 
• Bot Accuracy levels – Here you need to select the query matching the accuracy level in percentage.
• Set priority of bot and human agent – There are three options available to choose and set the priority of bot and human agent for seamless communication.
• Set trigger events – You can select triggers for the chatbots across different scenarios. Based on the selected trigger the chatbot will initiate conversation.

Note: REVE Chat offers four main triggers with sub-options. Based on the trigger you select, the chatbot will initiate a conversation.

1. Trigger the bot for – (i) All (ii) New Visitors (iii) Returning Visitors

• When you choose (i) All, the bot fires to all your website visitors & customers without any specific conditions like new visitors, repeat visitors, browsing duration of the visitor, etc. 
• When you select (ii) New Visitors, the bot will fire triggers only to the new visitors to your website by sending customized greeting messages like: “How may I help you?”
• On choosing the (iii) Returning Visitors, the bot will fire triggers to the same or repeat customer by saying “Good to see you! How may I help you today?”

1. Trigger the BOT for clicking on the chat banner – The bot will fire triggers, only when the visitor clicks on the chat widget. 
2. Trigger BOT based on URL – The bot will fire triggers when any visitors move on to a specific page on your website then the trigger will be fired if you choose this one. For e.g. Features or the Product page.
3. Trigger BOT based on browsing duration – When the visitor stays on your website for a specific time then the trigger will be fired on choosing this trigger event. Suppose your visitor stays more than 3 seconds on the pricing page, the bot will trigger “ Get 10% discount off on today’s sign up”.

3. After saving the bot profile the next step is the ‘Bot Builder’. It is a relatively important step as you define the conversational flow of the bot with different nodes’ actions, sentiments, and fallback options.

The first section is the ‘Path’ which gives a visual flow of the chat. Here you set the fallback options at the identified points, repeated query nodes, and sentiment nodes.

The center section is the ‘Responses’ that helps in setting on the desired answers for the commonly asked queries.

The third part is the ‘Actions’, which are combined in designing the conversational flow for your FAQ bot. You need to choose the required actions that can make your bot flow conversational and productive.

REVE Chat offers a list of actions described below:

• Text message – It helps in starting a conversation with your website visitors & customers with customized text messages or welcome greetings like “Good morning! How may I help you?”
• Question – You can set questions to collect your visitor’s information like – “Please share your email id and phone number.”
• Carousel -By choosing this action you can showcase your latest products in carousel ads format to the website visitors and customers for hassle-free navigation. 
• GO TO – It helps to escalate chat conversations in the right direction by connecting with the next step of bot flow as defined.
• Video – You can add a video chat step in your bot flow so that users can have face-to-face chat with the agent.
• API Plugin – You can connect to any systems via this action and collect the visitor login information.
• Image – Easily insert an image link for visual engagement of your website visitors & customers to guide them in their buying journey.
• Quick Reply – It helps in offering multiple options to your visitors to choose and move across bot conversations.
• Button – You can provide comprehensive details about your products and services to your visitors to help them in making quick decisions. 
• Audio – This action allows you to display an audio message to the customers during the bot conversations. You can set a URL of an audio clip or upload an audio clip.
• Chat Transfer – You can quickly connect your visitors with the right support reps when the bot is unable to understand the user request. 
• Email – Choosing this action helps to provide your customers an alternative channel to connect with you as a fallback option when the bot is not able to answer the query. 
• Appointment – It helps businesses in scheduling meetings and appointments in a smooth and hassle-free experience for customers.

4. After designing the bot flow, it is crucial to test the flow. On clicking ‘Test it Out’, a test website opens & the chat banner opens with a trigger message.

5. The next step is the ‘Intent Sample. Here you can create new nodes for creating intent samples.

• You have to enter sample queries having the same meaning that covers customer intention while interacting with the bot.
• When you enter the sample queries, a list is created that is used by the bot to handle customer service conversations.

6. The next step is sentiment analysis. Sentiment analysis helps in developing the bot’s emotional intelligence. Here you can ‘Add the offensive words and emojis’ for your bot to identify the user’s sentiment during the conversation.

7. The next step is setting the ‘Synonyms’. 

It allows you to configure synonyms from Bot Builder. If the customer uses jargon in his query, the system will understand all the synonyms of it and can use it to match with the corresponding knowledge base.

8. The next step is setting the ‘Fallback’ options.

Defining the fallback scenarios is an important part of designing chatbots. If your bot is not capable of fulfilling the user requests, it is not an ideal fit for those scenarios.

9. Now in the next step, you have to create FAQs. 

Creating FAQs helps your bot to learn about your business. It trains the bot to identify the user intent and have effective communication with the users. Here is how you can create FAQS to enhance the user experience.

10. The final step is training your FAQ bot. 

Chatbot training allows users to export bot training data in bulk in excel format. The users can edit the training data in the excel file and import it back in order to update existing bot training data.

And there you are. Your FAQ chatbot is ready to be deployed and manage customer conversations.